Thursday, February 17, 2011

Uncertainty or Definitiveness in Climate Science?

Two new studies, published in Nature (subscription required), find a link between anthropogenic greenhouse gas emissions and extreme weather events. At least, that's what I keep hearing. But something else is happening here, beyond the humans-are-changing-the-climate narrative. Behind the boilerplate conclusions offered by the two studies, and in the context of climate-fueled extreme weather research in general, is an interesting discussion of the value and validity of such specific scientific findings to policymakers and the general public.

As The New York Times reports:

In the first major paper of its kind, the researchers used elaborate computer programs that simulate the climate to analyze whether the rise in severe rainstorms, heavy snowfalls and similar events could be explained by natural variability in the atmosphere. They found that it could not, and that the increase made sense only when the computers factored in the effects of greenhouse gases released by human activities like the burning of fossil fuels.
The two studies each used climate modeling to predict the likelihood of observing discrete precipitation levels that, according to their findings, were impossible without the contribution of human-influenced atmospheric moisture content. The first study (Min, Zhang, Zwiers, and Hegerl) analyzed global rainfall over a fifty-year period (1951-1999) and used climate simulations to calculate the effect of increased GHG levels in the atmosphere on precipitation. The second (Pall, Aina, Stone, Stott, Nozawa, Hilberts, Lohmann, and Allen) established causal relationship, using climate simulations, between emissions and a unique extreme weather event (specifically, a flood in England in 2000).

The context for the two respective papers is important. The conclusion offered by the Min et al. indicates that climate change is both happening and having an effect on macro-scale phenomena, specifically global precipitation preceding extreme weather events. The paper by Pall and colleagues ostensibly asserts that we can identify a climate "fingerprint" in specific extreme weather events, indicting humankind for its increasing frequency.

Joe Romm at Climate Progress used the evidence put forward by these two studies in saying that "key weather events are becoming more extreme — especially deluges, heat waves, and droughts — as climate scientists have long predicted they would if atmospheric concentrations of greenhouse gases kept rising." He is referring here to the conventional wisdom in climate change research that, while we anticipate increasing and intensifying climatological extremes associated with anthropogenic emissions, no single weather event can be attributed to the effects of climate change. Climate Progress (which, incidentally, censored me just last week) seems to have put this conventional wisdom to rest.

Not so fast, say others. Over at his blog, Roger Pielke Jr. provides the following contribution to this discussion:
I have been asked by many people whether these papers mean that we can now attribute some fraction of the global trend in disaster losses to greenhouse gas emissions, or even recent disasters such as in Pakistan and Australia.

I hate to pour cold water on a really good media frenzy, but the answer is "no." Neither paper actually discusses global trends in disasters (one doesn't even discuss floods) or even individual events beyond a single flood event in the UK in 2000.
Pielke notes that the Min et al. paper discusses precipitation only, omitting any data on streamflow or damage. On this, he says,
Precipitation is to flood damage as wind is to windstorm damage. It is not enough to say that it has become windier to make a connection to increased windstorm damage -- you need to show a specific increase in those specific wind events that actually cause damage. There are a lot of days that could be windier with no increase in damage; the same goes for precipitation.
Further, Pielke criticizes the Pall et al. paper for drawing conclusions on extreme weather trends in a study with an overly narrow focus on a single weather event. The data, Pielke suggests, don't add up to such a compelling conclusion as has been reported.

Perhaps the most interesting dynamic at play here is what Pielke referred to as "a really good media frenzy." Andrew Revkin at DotEarth has probably the best compilation of opinion and fact surrounding this story, and he uses these two studies to discuss the interplay between scientists and journalists. Specifically, Revkin reminds us, climate science does not tend to lend itself to soundbites and easy conclusions:
As I wrote recently, there seems to be an inverse relationship between the definitiveness of an assertion and its credibility. This doesn’t mean that everything definitive is wrong (only Joe Romm could find a way to interpret it thus). It means that a reporter, or citizen, confronted with a flat statement on a tough issue would do well to dig a bit deeper.
Reporting on the issues, and making policy in response to them, becomes that much more difficult because of the simplistic diagnostics many of us seek. This is the key trouble with climate science, in my view. Science and journalism very often present us with definitive climate predictions, and when credible non-deniers like Pielke chime in with data-driven skepticism, the zeitgeist of climate change as a whole suffers in the public arena. The trouble with trying to solve the climate crisis by using climate disasters as harbingers is that we have yet to accurately predict in advance or attribute in retrospect any direct culpability to anthropogenic emissions in high-profile weather events. Thus, insofar as climate change is a technical challenge with a technical solution, it is extremely difficult to define the parameters which we seek to address.

These two studies, and the reporting on them, are examples of how messy the arena of public discourse can get without even inviting the flat-earthers who refuse to entertain the notion that climate change might just possibly be a bad thing. It is for this reason primarily that I (taking a cue from Revkin) like to frame the challenges we face as an energy quest, for which the technical challenges posed are far more simple to diagnose, even if they are just as daunting.

No comments: